How the Mind Works: The Dunning-Kruger Effect (Norman Costa)
A Facebook friend, Jose Del Solar, posted the folowing article on the Dunning-Kruger effect: Incompetent People Too Ignorant to Know It.
The Dunning-Kruger effect is fascinating, and I've written about it before. In short, people with little knowledge are unable to appreciate their intellectual deficit. Instead they view themselves as very knowledgeable. They overstate their own intellectual competence. On the other hand, people who are very knowledgeable, see themselves as far less accomplished.
Most people will pass through a low knowledge episode or short period of the D-K effect in the course of being educated, growing, and learning. In fact, we may do it multiple times, usually before we get out of high school.
I remember one of my own experiences in 10th grade biology. We had many classes on human blood and the circulatory system. A friend and I agreed that we learned a great deal of the subject matter, compared to our total ignorance at the start of 2nd year biology. We could not imagine that there was more to learn, because it seemed, to us, as comprehensive as we could imagine.
It may sound cliche, but as we learn more and more we appreciate how little we know about the larger subject matter areas. D-K folks never get that far. The D-K effect is evident in just about any subject matter area you can imagine.
What I have found so bothersome, and troubling, is the number of people in positions of power (corporate, military, politics, varied institutions) who cannot imagine that there own knowledge of the world is deficient. The two most obvious examples that come to mind are George W. Bush, and Sarah Palin.
Even when they are brought up to speed on an area where they were deficient, they still cannot see that there could be more to come. To you and I, this is obvious, and we can't get our minds around the fact that they are completely oblivious. Palin was never embarrassed by the fact that she thought Africa was a country, not a continent. However, she was pissed that others made her feel inferior because she was not in the know.
One of the post-hoc giveaways of the D-K effect is that the person who leaves a position of great power is unable to manifest a sense of humility at confronting and dealing with the things that demanded great intellectual effort and new learning. They don't express a sense of awe and respect for what others who preceded them had to deal with, nor any sympathy or compassion for the next person who will occupy the seat of power they just vacated. Again, Bush and Palin are two great examples.
There is another phenomenon that I've observed, and it may be a variation on the D-K effect, except that you find it in otherwise knowledgeable and smart people. This happens when confronted with a problem to be solved, or an area of new knowledge that is important to the present situation. They come to an solution, or an understanding, very quickly and then they hold onto that idea or position with enormous tenacity. One exceptional example was Alexander Haig. He was the former Supreme Commander of NATO forces, Whitehouse aide to President Nixon, and former Secretary of State. I've known a number of these people in executive positions in IBM Corp. This is not a matter of blind adherence to ideology. It's a very different animal.
It is always interesting to try to divine the thought process of political figures. Since this is a Presidential election year, it's a more salient matter. Rick Santorum, in my opinion, formed every significant thought he ever had by the time he was 21. Intellectual life, since that time, has been a process of seeing anything new as a variation on what he already knows. There is a prearranged slot on the shelf for anything that comes along, and no need to create any new slots or reconfigure the old ones.
Mitt Romney, in my opinion, is very much aware of his intellectual limitations and the availability of other view points on just about any issue. Newt Gingrich is harder to read. I am not sure if he displays a thought process that discounts new learning, and overvalues what he knows, or if he is just being confrontational, insulting, and controversial as a strategy, or as a character defect.
I have this mental image of showing the article I referenced, above, to someone who lacks knowledge yet believes they know just about all there is to know. After reading the article they look up at me and ask, "Yeah, so what's your point?"
Norm... good posting. As you indicate, it is not only at IBM that one can come across these D-K types. Having spent 20 years in big industry I have found them here and there. I call them the "one big idea" type. They get the "big idea", find some degree of success employing it, and never let it go. Why does Al Gore come to mind? He seems to be fixated on Global Warming and does not add new thought patterns to his position, even when falsified data comes to light. Note I am taking Global Warming as an example, not taking a position one way or the other.
Posted by: Louis DeFilippi | February 28, 2012 at 07:47 AM
Based solely on this post, the linked article by Wolchover, and the occasional references to D-K effect I typically encounter on political blogs, I think the phenomenon has become a figure for something else not really addressed by the research. It supplies a weapon for name-calling. Accusing somebody of suffering from the D-K effect is like calling him or her a numbskull, with the added advantage of being able to explain why he or she is a numbskull, namely, because he or she is too stupid to know how not to be one. But where does the explanation get us? Well, it teaches us we shouldn't rely on anybody to X who doesn't know how to X, who nevertheless thinks he or she can X just fine, and who consequently is unable to learn to X.
Including politicians in the mix here complicates things. Sure, Bush and Palin are numbskulls. But immune to embarrassment or humility? How are we supposed to know this? They're both politicians, eager for power and command. Even if they do feel embarrassment or humility, they are going to sublimate it and forge ahead with the task of self-promotion required to succeed as a politician. What, after all, is the measure of capacity here? Bush was POTUS, indisputably, which stands for something. Palin...I don't know.
My point here is that D and K focused on relatively narrowly definable "knowledge domains." Being POTUS or being a politician is not such an easily definable domain, unlike even the very weird example given in the article, judging jokes (see the graph). (They relied on "professional comedians" to rank the comic value of jokes? Really? That's an area of expertise? And it's exclusive to professional comedians? That's like asking mothers to judge the quality of home cooking.) The question explored isn't whether the test subjects lack knowledge altogether, whether they're completely inept buffoons like Santorum, but how they evaluate themselves in a more precise domain where the extent of their knowledge is uncertain. If I read these materials correctly, Alexander Haig, as smart as he was, may yet suffer from D-K effect. He might have been a bad judge of his own sense of humor, for instance. Accordingly, I wouldn't make D-K positive or negative the deal-breaker in a presidential election.
Posted by: Dean C. Rowan | February 28, 2012 at 03:07 PM
@ Dean:
What are you trying to do? Keep me honest and consistent? OK, but I'll get back after my sesame chicken. I haven't eaten all day.
Posted by: Norman Costa | February 28, 2012 at 03:48 PM
Nor had I when I commented, Norm. Now that I've had my chicken parmesan [sic] hero, let me hypothesize a Costa-Rowan effect, which identifies people with more than at least a marginal threshold of knowledge in a domain, who are generally well aware of their own informational and cognitive limitations, and who nevertheless exploit opportunities to stretch the truth, if only for the fun of it.
Posted by: Dean C. Rowan | February 28, 2012 at 05:24 PM
@ Dean:
Having trouble with the disappearing comment. Gotta try Ruchira's recipe for fixing the problem.
Posted by: Norman Costa | February 28, 2012 at 08:14 PM
I am severely afflicted by the Costa-Rowan effect. Which is why I began blogging.
Posted by: Ruchira | February 29, 2012 at 04:38 AM
I remember there was a wonderful online test of this, which I can't find any more. The idea was to ask people to answer a hundred questions, and for each one to provide a subjective degree of confidence. Then that subjective confidence was compared to actual facts. The person perfectly aware of his limitations would always get those subjective confidences right. So, 65% of questions he answered with 65% confidence would in fact be answered correctly. The result was presented as a curve plotted as actual percentage vs subjective confidence, with the set of ideal responses all mapping onto the right-diagonal line. Does anyone remember this and know where to find the test?
Posted by: prasad | February 29, 2012 at 07:37 AM
The test is also a direct probe of the DK effect of course - compare the deviation of the curve above to the actual fraction of questions answered correctly.
Posted by: prasad | February 29, 2012 at 07:41 AM
Dean:
Wikipedia has as good a description of the Dunning-Kruger effect as any:
"The Dunning–Kruger effect is a cognitive bias in which the unskilled suffer from illusory superiority, mistakenly rating their ability much higher than average. This bias is attributed to a metacognitive inability of the unskilled to recognize their mistakes.[1]
"Actual competence may weaken self-confidence, as competent individuals may falsely assume that others have an equivalent understanding. As Kruger and Dunning conclude, "the miscalibration of the incompetent stems from an error about the self, whereas the miscalibration of the highly competent stems from an error about others" (p. 1127).[2]"
The title of the article, above, is typical of magazine reporting. I deliberately kept it out of my own title. You've already noted that some of the research in that article is highly limited in generalization. Some of it is highly controlled, even contrived, to test the D-K Effect in as pure a form as possible. Fortunately, the research on D-K is over a dozen years old and stands up to a lot of variability in content area and experimental set up.
It's important to note that it is a cognitive deficit, not a personality disorder - an error of self vs. an error about others. I mentioned my error of self in my posting. Since then, I identify more with the errors about others.
It's always risky to extend research results to describing actual events or personalities, as I've done here. I've used the phrase, "in my opinion," several times. My personal view is that Sarah Palin is the easiest to justify as an example of error of self in the D-K model. Bush is a bit more complex. For example, in my view, his thought process is consistent with that of a dry drunk. He may have been abstinent for many years, but I am not aware that he did anything to correct the problems associated with an addictive or alcoholic personality. He drank a lot of non-alcoholic beer in his sobriety. This is not a smart thing to do, since non-alcoholic beer is NOT alcohol free.
Your comments bring up another issue that has plagued psychological research from its earliest history. It not enough to say that individuals differ on various traits. Some attach social or moral values to the extremities of the distribution. People differ on various scales of 'creativity.' The high end is thought of as GOOD, and the low end is BAD. This has caused some researchers, over the years, to refer to such constructs as personal styles, rather than personal traits. The D-K Effect is a tricky trait to deal with. I think it is safe to say that D-K is unique to a given knowledge domain in particular people. I haven't seen anything that describes D-K as a generalized cognitive style or personal trait.
Posted by: Norman Costa | February 29, 2012 at 08:41 AM
For what it's worth, I took the 'at sign' (@) off my comment, above, and it appeared at shown. Go figure.
Posted by: Norman Costa | February 29, 2012 at 12:44 PM
No luck finding the online test Prasad recalls, but I have stumbled across this remarkable piece of journalism out of the NYT only two years ago. I don't have time to read it all now, but some of the juicy parts (a pun, you'll see) are at the outset, focused on the D-K Effect, a little later in a bit about President Wilson's stroke, and at the end, where Dunning contributes a Venn diagram depicting the relationships of our varieties of human ignorance. This article confirms my sense that the D-K Effect gets treated almost moralistically. Even the discussion by Dunning himself shuttles from neutral ascriptions ("incompetence") to emotionally or morally charged ones ("stupidity," "idiots"). Yet ironically, as Morris mentions in a footnote, the acknowledgement of our own limitations can be a feature of a kind of wisdom (Socrates or even Donald Rumsfeld).
There are several other questions posed by this issue. For one, there seems to be a tacit assumption that the incompetence and its reciprocal competence are specific to individuals. I don't know how to X, but somebody out there does, and this circumstance makes my inability to know that I don't know how to X seem ridiculous or unfortunate. But there are a lot of widely known gaps in knowledge, such as a cure for the common cold. Are there also widely unknown unknowns, and do "we" therefore suffer a kind of group D-K Effect? Dunning seems to think so. "Evolution just makes sure we're not blithering idiots." Also, the quest for knowledge can easily transmute into a quest for mastery, which is different. Do we really expect people to know themselves by knowing (by cataloging or compiling in an ambitious Diderot's Encyclopedie-like compendium of universal ignorance) what (or merely that) we don't know?
I dunno.
Posted by: Dean C. Rowan | February 29, 2012 at 01:58 PM
Dean:
Good article. Didn't read it all yet, but it looks great, so far.
Dunning is a social psychologist and focuses his research interests on decision-making. While I don't know much about Dunning, himself, I have a few observations about social psych research. This field has a long history of experimental 'laboratory' research. This is compared to field, or naturalistic, research. Observing kids and their social interactions in a playground is field research. Doing a highly controlled decision-making experiment using a very contrived, or artificial, task is 'lab' research. You commented on this earlier. 'Laboratory' experiments tend not to look like anything you would encounter in the real world. The real distinction between 'lab' and 'field research' is that in 'lab' research you have maximum control over as many experimental variables as possible. The definition has nothing to do with the fact that the experiments are conducted in university laboratories.
All things being equal, results from field research are more generalizable to more situations in the real world, but there is much more error variance in predictions. 'Lab' are less generalizable but have less error variance in predicting.
The psychologist, Herbert Simon (1916 - 2001), was one of the pioneers in research in decision-making. Many of his research experiments used the template of the "Towers of Hanoi" puzzle - talk about your contrived experimental setting. He is the only psychologist ever awarded the Nobel Prize, in Economics of all areas, for his work on decision-making in economics.
Posted by: Norman Costa | February 29, 2012 at 02:27 PM
The Costa-Rowan Effect:
There is a variation on this which we can call the Costa-Rowan^2 Effect. That's when we labor under the delusion that the rest of the world is interested in what we have to say.
Then there's the Ruchira Effect. That's when someone who manifests the Costa-Rowan Effect starts a blog, and unwittingly creates a venue for everyone else that manifests the Costa-Rowan Effect.
Posted by: Norman Costa | February 29, 2012 at 05:30 PM
A tangent: the distinction between laboratory and field research is much like that between Boolean and "natural language" (i.e., algorithm-driven, where we don't know the underlying algorithm, e.g., Google) searching. With the former, we know exactly why we obtained our results, but we are less certain that we've retrieved a large number of the relevant ones. The latter offer the prospect of more relevant, richer results, but at the price of not being able to control how we retrieve them. Put another way, laboratory research, like Boolean searching, invites a version of the "between the sheets" game where participants recite a movie or song title or lyric, or a passage from the Bible, what have you, followed by "...between the sheets." Thus, "2001: A Space Odyssey Between the Sheets" or "Thou shalt not commit adultery between the sheets." Laboratory research involves a similar rule: "We found such-and-such...where we controlled for variables X, Y, and Z." And Boolean searching similarly invites a qualification, often left unstated: "I found everything pertaining to cats and dogs...but not felines and canines." If we're being rigorously forthcoming about the limitations of our (re)search, we'll remember to add the "between the sheets" part.
Posted by: Dean C. Rowan | February 29, 2012 at 05:35 PM
Both the Costa-Rowan Effect^2 and the Ruchira Effect, of course, are species of the more widespread network effect, without which the 'net and various mobile technologies would be more or less useful and entertaining, depending on your point of view.
Posted by: Dean C. Rowan | February 29, 2012 at 05:38 PM
Hmmm. ...using the Towers of Hanoi puzzle to make economic decisions between the sheets.
Posted by: Norman Costa | February 29, 2012 at 07:14 PM
A little unorthodox, but I differ with those who see the DK effect as a cognitive deficit quite separate from personality. There are personalities that switch into action mode precipitately. THEY however need to feel they are prepared. Often, they will say it comes down to a gut-check, which is their way of saying they don't need or want the kind of prep you need to make a decision, or just to get a good fast take on something difficult and important. I believe the lens through which they see things is not the limited and perhaps inaccurate info they possess, which they are sticking to, and which they run everything by, but the disregard they feel for people who know enough to have informed opinions about the same subjects they merely flash on. I don't see this as a cognitive deficit -- if it were, would it be accompanied by arrogance and disregard? The ignorance that knows not of itself is found in some perfectly fine people who do not want to put others down, after all.
Posted by: Elatia Harris | March 02, 2012 at 10:58 PM
Elatia,
The D-K Effect, as described by D and K themselves, is clearly a property of cognitive functioning (more or less.) However, the predilection toward the 'gut check' as a dismissal of the utility, or even necessity, of more valid information for decision-making may be a personality issue. I go back to my IBM days when I saw this in action time and time again. The big difference with the D-K Effect is that the dismissive 'gut check' was actually a manifestation of feelings of insecurity, and hostility toward those who were in-the-know, and had better skills and more information. You rarely saw this in a peer to peer interaction. Rather, the 'gut-checker' was in a higher power position, and reacted in fear that the more knowledgeable (or one who favored getting more information) could devalue the judgement of the one with more power. It's the 'old bull of the woods' general manager being threatened by the young MBA. Any arrogance I saw was just another expression of feelings of inadequacy and resentment. This is more a personality matter.
Sarah Palin, I think, is still a good illustration of the difference. Clearly, the country of Africa thing was more a cognitive deficit. However, getting angry because the knowledgeable folks made her feel insecure and embarrassed speaks to her emotional immaturity (personality.) Another example had to do with Tina Fey's skit and her line, "And I can see Russia from my house." I was listening to an interview with Palin a couple of months after the SNL broadcast. Palin understood that her appearance on SNL was equivalent to agreeing to be dropped into the dunk tank. For that she had to be a good sport, and she was. In the interview, however, she made reference to Tina Feys line and showed that she did not understand that the line and the skit were parody. She said to the interviewer, in a resigned and slightly despairing tone, "But, you know, I never really said that."
To the late night comics, Sarah was the gift of cognitive deficits that kept on giving.
Posted by: Norman Costa | March 02, 2012 at 11:55 PM
Thanks, Norman. As you point out, even smart people may overestimate their knowledge about a certain field -- esp if they are students who just took a survey course. Experts in the same field may have seen into it so deeply they are genuinely humbled by all there is yet to learn. Bernard Berenson remarked -- "Every serious subject is infinite." But it's hard for me to swallow that a near know-nothing with an attitude, and with the seeming confidence and arrogance that can cover deep insecurity, has nothing going on a but a cognitive deficit. That seems to me like saying that real narcissists are merely stuck up.
Posted by: Elatia Harris | March 04, 2012 at 12:34 PM
What about the reverse of the Dunning-Kruger effect, where people routinely underestimate their abilities rather than overestimate them, even if they exhibit low skill levels to start with? As this article shows, it may also be a culture-specific thing.
There is an old Tamil proverb which translates to "What has been learned is a handful, what hasn't been learned is as huge as the ocean". This mindset may be more frequent in Asia than in the US.
Also, this raises the question, if we know that a person is manifesting the D-K effect, can we blame them for their own inability to not perceive their lack of competence? How worthy would such a person be of our scorn, or are we setting ourselves up to be the next targets where we feel remorseful at having unnecessarily judged someone as incompetent when we weren't aware of our own?
Posted by: Sujatha | March 04, 2012 at 12:43 PM
Sujatha, can one be scornful of people with cognitive deficits? I think it's okay to be scornful of people who are arrogant, uninformed and hellbent on staying that way -- but I believe that's different from a cognitive deficit. If someone is ignorant in the sense they are uneducated, or stupid in the sense their ability to learn is very limited -- well? And what happens to DKE-afflicted people when they finally discover, or are forced to admit, they really don't know jack about the subject they were so complacently in control of?
That's a beautiful proverb. Real humility and modesty is rare, perhaps because the marketplace reads it as failure of competence and confidence, instead of a balanced notion of what one person may be truly on top of in one lifetime.
Posted by: Elatia Harris | March 04, 2012 at 04:35 PM
Elatia,
What you describe is certainly a manifestation of personality deficit, with or without any cognitive deficit. Dunning and Kruger, however, have focused only on the cognitive issues, so their work can't really enlighten us on any attending personality dysfunctions - though they may be legion.
I don't think we need the DKE to understand insecurity and byproducts like arrogance, narcissism. and all sorts of anxieties. Personally, I don't think anyone has said it better than Alfred Adler and the idea of over-compensation for feelings of inferiority [inadequacy, insecurity, and the like.]
My discussion of the DKE is a technical one, and circumscribed at that. If you find a person who is low informed, but overrates their competence, you won't find it generalized to all subject matter areas. Someone with great insecurities who manifests with various ways of overcompensation, can be found to be arrogant, superior, dismissive, demeaning, and so on in many areas of life - including subject areas where they are virtually ignorant.
Sujatha's and your discussion remind me of the Nobel physicist, Steven Weinberg. Early in his doctoral level career, he believed the ultimate answers to questions of physics would yield to determined scientific research. His only misgiving was he might not live long enough to see the answers to the great questions about Nature. Eventually, he came to believe that there are no ultimate answers to reality, no matter how long you live. He accepted it as the nature of the human condition - we will never know. For Weinberg, the study of quantum physics led him to a crisis of faith in himself and science. For Weinberg, it was a profound and disturbing awareness. Eventually, he accepted the nature of the human condition, as he understood it. Once you study quantum physics, he said, you are changed forever and there is no going back.
Posted by: Norman Costa | March 04, 2012 at 07:50 PM
So, would the cognitive deficit exist in neurotypical people who did not have the personality type we have been discussing? Or would the personality type form around the cognitive deficit?
Posted by: Elatia Harris | March 04, 2012 at 09:21 PM
Elatia,
Q. "So, would the cognitive deficit exist in neurotypical people who did not have the personality type we have been discussing?"
A. YES, as I understand the work of D and K.
Q. "Or would the personality type form around the cognitive deficit?"
A. NO, as I understand the work of D and K.
However, I would like to hedge a little here. D&K were experimental social psychologists and not child, developmental, abnormal, or personality mavens. They were interested in decision-making. A clinical researcher, personality researcher, or researcher in abnormal psychology would approach this issue very differently. They might feel that the personality type would form around the cognitive deficit - at least to some degree. Keep in mind that the kind of lab research that social psychologists conduct is designed to eliminate as many variables as possible. Also, they are NOT interested in the contribution of the DKE to personality formation, nor the contribution of personality to the DKE.
Bottom line: Others, more knowledgeable than I, may be familiar with research to support your observations and interpretations. The DKE, as formulated by D&E, do not lend any support.
Posted by: Norman Costa | March 04, 2012 at 09:48 PM
Dunning-Kruger, a presidential ticket for 2012?
Posted by: Dean C. Rowan | March 05, 2012 at 10:50 AM
Dean,
Thanks for the link. This is one of those person-in-the-street generalizations from a news clip about results from psychological research. I read stuff like this and I start to regress - I want to scream, stamp my feet, pull my hair out, and bang my head against the wall. I'm gonna make a pot of tea.
Posted by: Norman Costa | March 05, 2012 at 02:40 PM
Thank you, Norman. We'll have to wait a bit longer for all the blind men to agree about the elephant, I guess. Meanwhile, Dean -- Dunning-Kruger 2012!!!! W00t!
Posted by: Elatia Harris | March 05, 2012 at 05:28 PM
OK, folks. Just when you thought it was all over - I just finished watching the HBO movie, "Game Change." Julianne Moore played Sarah Palin in the 2008 Presidential Election. Anyone else watch it and care to comment?
By the way, Julianne More was superb.
Posted by: Norman Costa | March 11, 2012 at 11:03 PM
I didn't watch it, mostly because I have no access to television. But I probably wouldn't have watched it, anyway. I'm reminded of Jean Baudrillard's audacious proclamation, "The Gulf War Did Not Take Place," the title of a book published in English in 1995. The relationship of McCain-Palin '08 to contemporary SNL parodies of Palin to a movie four years later that, I presume, incorporates the history of the campaign and its popular reception via SNL, is kaleidoscopically mind-boggling to me. Is the movie an account of a campaign that barely took place? SNL's mockery commanded at least as much attention as the "real" figures and events involved, and caused a feedback loop, the soundtrack to Palin's caricatured enactment of herself. Do we have to watch Tina Fey mock Julianne Moore portraying Sarah Palin? Or Fey mimicking Moore mocking Palin? Or Moore portraying Palin mocking Fey?
Your response, Norm, to the person-in-the-street "news" is exactly mine when I read a good chunk of pop science literature, or stuff purporting to prove the Internet makes us stupid, this latter being a kind of self-fulfilling literature.
Posted by: Dean C. Rowan | March 12, 2012 at 01:37 PM
Dean,
Thanks for the reference to Baudrillard. I was unaware (as for many things) and I checked it out. I always like entertaining speculative views and surreal reversals, if only to compare it to reality. I found his "It's not going to happen," "It's not happening," and "It didn't happen," as templates for a different opinion (very different) on something momentous. I guess some called it instant revisionist history, but it strikes be as, 'now let's look at this in a very different way." The mind needs an astringent, once in a while.
"Game Change" is a movie about politics like I have never seen before. If Sarah Palin never existed, there would be no need to make such a political movie. The movie is about Sarah Palin and so much more. I do not like Sarah Palin, I do not want to be her friend, or her mine. She is, fundamentally, a dishonest person. By that, I mean she lies left and right to protect her view of herself. But the fact that she ended the campaign standing on her feet, and walking under her own power, was extraordinary. If one goes to see the film to watch her tell the world she's an ignoramus, or to cheer her on against those educated elitists, you will forget all that and watch the movie you didn't know was there. And the acting and writing and directing were all first rate.
As far as the movie went, Ed Harris did a great job on an understated and underpowered McCain. Woody Harrellson played a character that you wanted to shoot, and wanted to love. The actress who played Sarah's early speech writer made you wince and grimace and resent right along with her. Julianne Moore must get an Oscar. That is all I can say.
Posted by: Norman Costa | March 12, 2012 at 04:49 PM